<<

. 33
( 62 .)



>>

to constitute part of new, locationally wide abilities, which are then used
to interact with other mediators, leading to further locationally wide abil-
ities, and so on. Once such wide, mediational abilities are established,
they can then be transformed and internalized by individuals and thus
decontextualized from their initial context of acquisition.
A part of the child™s environment that serves as a crucial mediator
across a range of contexts is other people. Here some work within the
Thinking Through and Beyond the Body
202

recent “theory of X” approach in developmental psychology shows con-
vergence with some of the conclusions central to the Vygotskyan tradition.
The work that I shall discuss in the remainder of this section focuses on
explanation and how a grasp of it develops in children over time.17
A striking paradox about explanation and how it is cognized turns on a
triad of claims. First, explanation is a crucial and ubiquitous part of much
of everyday life. Second, its ubiquity in our day-to-day dealings with one
another seems to re¬‚ect the various “theories” that we have about corre-
sponding domains. Yet third, the level at which we are able both to offer
and understand explanations is amazingly shallow. What I mean by this
is simply that explanations typically stop or bottom out surprisingly early
on. Consider an intuitive example from the developmental psychologist
Frank Keil. Although almost everyone who owns a car can give some sort
of explanation as to why their car starts (or doesn™t) when the key is placed
in the ignition and turned, few of us are able to respond with any depth to
even the next few follow-up “why” or “how” questions. The shallowness in
this case is the norm: We rarely have ready access to explanations of any
depth for all sorts of phenomena for which we are able to offer some sort
of explanation. Indeed, we often carry with us what Leonid Rozenblit and
Frank Keil call an illusion of explanatory depth until we are faced with the ac-
tual task of explanation. Thus, people frequently seem to think they have
vivid, fully mechanistic models of how something works or how it got the
way it did. But when forced to state explicitly that mechanism as an expla-
nation, their own intuitions of explanatory competence are shattered.
Although overcon¬dence in one™s abilities has been experimentally
documented in a range of domains, this experimental work has been
recently extended to apply to explanation in particular. For example,
Rozenblit and Keil asked college students whether they know how vari-
ous familiar devices work, such as ¬‚ush toilets, piano keys, and zippers.
Subjects were given a list of forty-eight familiar items, and asked to use
a seven-point scale to rate how well they understood how each worked.
They were then given a four-item test list and asked to produce a detailed,
step-by-step causal explanation for each item on this list, followed by a
series of reratings of how well they understood each of these items follow-
ing their attempt to produce an explanation and their reading of expert
explanations. Rozenblit and Keil also ran versions of the experiment with
facts, narratives, procedures, and natural phenomena.
The general ¬nding of this series of studies supports the original claim
that Keil and I made about the shallows of explanation: There is an illu-
sion of explanatory depth in ordinary people™s understanding of familiar
devices. Many subjects assert or imply that they have a complete and fully
The Embedded Mind and Cognition 203

worked out understanding such that they could explain all the neces-
sary steps in any process involving the object. Yet when asked for such
explanations, a large percentage of these participants show striking in-
abilities to put together a coherent explanation, missing not just a few
arbitrary details, but critical causal mechanisms. Until they attempt such
explanations, they are often under the illusion that they have a com-
plete, “clockworks,” vivid understanding. The in¬‚ation in the estimate of
one™s ability to provide an explanation seems not to be general, since the
pattern of results obtained in the device condition was not repeated in
the facts, narrative, or procedure conditions, although it was repeated in
the natural phenomena condition. Thus, there seems to be something
about our knowledge of both familiar devices and natural phenomena
that creates this illusion of explanatory depth. Rozenblit and Keil suggest
that these domains promote the illusion because they are sites at which
a series of confusions converge. For example, people mistake something
like the perceptual vividness of familiar devices and natural phenomena
for the richness of their own internal representations, and their under-
standing of these at their functional or operational levels with that of the
underlying causal mechanics.
Apart from the illusion of explanatory depth itself, what is striking
and relevant for developmental paradigms that emphasize the nature
of underlying theories that guide cognitive development is the lack of
explanatory depth itself. Missing in adults seems to be the sort of the-
oretical understanding of the underlying workings that anything more
than a skeletal theory in the relevant domain would surely provide. The
shallows of explanation thus pose a prima facie problem for accounts of
cognitive development that ascribe to the child a theory of the relevant
domain, which is then added to or modi¬ed through development. This
is because the shallows of explanation we observe in adults seem to be
due to a sort of theoretical abyss. The gap between the limited expla-
nations that suf¬ce for the purposes at hand and those that we believe
ourselves to have is due ultimately to a lack of corresponding, detailed
theoretical knowledge that would allow us to provide more satisfying,
detailed explanations. That is, the shallows of explanation are not due
simply to contextual or pragmatic features of the practice of explanation,
or from general processing or access abilities. Rather, it stems from some-
thing central to the “child™s theory” view of cognitive development: the
absence of detailed theories themselves.
This implies that the problem in the triad generating the paradox con-
cerning explanation (ubiquity-theory-shallows) lies in the second claim:
that explanations re¬‚ect an individual™s theoretically rich understanding
Thinking Through and Beyond the Body
204

of some phenomenon. If explanation is theoretically shallow, how can it
also be ubiquitous?
What makes explanation both ubiquitous and shallow is a certain divi-
sion of cognitive labor, one which makes detailed theoretical knowledge
in the head of each individual unnecessary, a division akin to the so-
cial division of labor central to Putnam-Burge externalism. We rely on
knowledge in others extensively in our explanatory endeavors (as in our
linguistic endeavors), and we rely on the assumption of knowledge in oth-
ers to give us a sense of explanatory insight. Everyday folk know enough
about the “nominal essences” of the things that they interact with on a
regular basis in order to be able to offer relatively shallow explanations for
their behavior. But there are also experts who have either the within-level
or across-levels knowledge that the folk typically lack, and who are in a
position to offer explanations with more depth. Although we are faced, as
individuals, with the theoretical abyss as the norm, the theoretical knowl-
edge that we lack exists somewhere, just not in our heads. This is to say
that explanation and the theories that underwrite their depth, are wide:
They do not supervene on an individual™s intrinsic, physical properties.
This reliance on the knowledge of others in explanatory understand-
ing seems to be a feature that even preschool children recognize, as re-
cent work by Donna Lutz and Frank Keil indicates. For example, a child
might be told that Bill knows all about why two magnets, if turned the
right way, stick together; and that John knows all about why a china lamp
breaks into pieces if it falls off a table. The child is then asked who knows
more about why television screens get all fuzzy sometimes during thun-
derstorms. Even preschoolers will cluster explanations about electricity
and magnetism together to a greater extent than either of those explana-
tion types with mechanics. There is no doubt that they are in nearly full
ignorance of any speci¬c mechanisms, yet as early as the age of four they
have some sense of how some explanations are more likely to be related
in the minds of experts. (In this particular example they may be keying
into notions of invisible forces and action at a distance.) As Lutz and Keil
report, even three-year-olds have a sense of how knowledge is clustered
in the minds of at least some experts, although this knowledge is far less
secure and extensive than it is at four.
One may think that this simply reveals more complexity to the internal
structure of the child™s mind than one might have initially suspected. In-
deed, die-hard strong nativists and individualists might even see it as pro-
viding evidence for another module, that for deference to others! But this
would be to miss the signi¬cance of these results and their relationship to
The Embedded Mind and Cognition 205

those concerning the illusion of explanatory depth. What both indicate
is not simply that children are built to rely on the knowledge of others but
that their actual reliance limits the extent to which internally rich struc-
tures are needed to understand the world around them and function
effectively in it.
This cognitive division of labor and the relatively impoverished, in-
ternal cognitive structures that go with it are instances of more general
features of cognition. As we rely on other people, so too do we make
use of information in the world more generally. And as we have at least a
skeletal understanding of what others know (and who knows what), so too
do we have “modes of construal” of the various ways in which properties
are causally clustered and distributed in the world. The general point is
that throughout much of development, and long before formal school-
ing, a set of framework explanatory schemata are at work and seem to
be essential for further theory growth and conceptual change. This sug-
gests the basis for some rapprochement between the new developmental
psychology and a mediational view of cognitive development.18
Vygotsky himself had a speci¬c, general view of how mediational abili-
ties developed. All mediational abilities, he claimed, begin as interpsycho-
logical abilities, abilities that are developed and manifested only in social
relations between individuals. This is a consequence of his “general ge-
netic law of cultural development,” which says that

[a]ny function in the child™s cultural development appears twice, or on two planes.
First it appears on the social plane, and then on the psychological plane. First it
appears between people as an interpsychological category, and then within the
child as an intrapsychological category.19

While this is plausible for communicative abilities, it is incompatible with
the results of what I have been calling the new developmental psychol-
ogy, even Keil™s modes of construal tempering of it, which indicate the
theoretical and conceptual sophistication that young children bring to
cognitive tasks “on the social plane.”
In any case, Vygotsky™s “law” seems less plausible for the full range of
forms that even mediational cognition can take. For example, many of the
mediational devices that have been culturally developed, such as maps,
signs, and even numerical systems, can be and are used by individuals
from the outset, particularly once they have already acquired related
psychological abilities. They do not appear to have the double existence “
¬rst on the social plane, and then on the psychological plane “ that
Vygotsky™s law attributes to all higher cognitive processes. But even if some
Thinking Through and Beyond the Body
206

mediational cognition is not tied as closely to a social context as Vygotsky
seems to have thought, the more general point that mediational cognition
is ontogenetically prior to many of the higher cognitive processes that are
the focus of individualistic psychology gives us reason to rethink what it
is we are discovering when we discover the “child™s theory of X.”
I want to conclude the substantive part of this chapter with some
thoughts about externalist psychology and the child™s theory of X, where
X = mind.


6 folk psychology and the theory of mind
We are mindreaders. The explosion of work over the last twenty years
in both cognitive development and primatology exploring the develop-
mental and evolutionary origins of this ability has largely construed the
capacity itself as a theory of mind, a theory that attributes folk psycholog-
ical states to agents, and that allows one to predict and explain an agent™s
behavior in terms of the relationships between those states, perception,
and behavior. Folk psychology itself forms a core part of many areas within
psychology, including work in social cognition, group dynamics, and deci-
sion making. Philosophical discussions of folk psychology concentrating
on the relationship between folk psychology and a truly scienti¬c psy-
chology have sometimes implied that such work could form no part of a
truly scienti¬c psychology. Yet such eliminativist conclusions have not gar-
nered widespread support either in philosophy or in psychology. Here I
focus my discussion of our mindreading abilities on the end state of these
ontogenetic and phylogenetic processes, the folk psychology that we end
up sharing and relying on in everyday life.
Folk psychology has played a prominent role in the debate between
individualists and externalists, beginning with the widely accepted conclu-
sion from the Putnam-Burge thought experiments that folk psychological
states are externalist. In Part Two, I argued that we should see this as en-
tailing that folk psychological states have a wide realization, and that folk
psychology itself is a type of wide cognitive system, one that involves the
social relations between individuals. But what of the underlying capacity
or disposition that each of us has to construct such a folk psychology?
The large and growing literature on our mindreading abilities has, by
and large, treated this capacity as an individualistic system, a theory of
mind module or the child™s theory of mind.20
In discussing this I want to distinguish between our bare-bones folk
psychology, belief-desire psychology, and a richer conception of folk
The Embedded Mind and Cognition 207

psychology, one that includes the full range of psychological states, such as
emotions (anger, elation, fear), moods (restless, horny, inattentive), and
sensations (pain, experiencing red, tickling). I shall refer to this richer
conception as full-blown folk psychology. Consider, ¬rst, bare-bones folk
psychology.
It is plausible to think that the capacity that normal human adults have
to ascribe belief and desire to one another is both locationally narrow and
taxonomically wide. It is locationally narrow because the realization of the
capacity is purely internal to the individual who has the capacity. But it is
taxonomically wide because beliefs and desires are individuated, in part,
by their intentional content, that is, what they are about, and such content
is wide. This is so whether one thinks that this ascriptive ability operates
via a theory or via acts of imaginative simulation. Bare-bones folk psychol-
ogy admits of what we might think of as an intellectualist construal, one
in which the ability itself is purely internal to the cognitive agent, and
I suspect that this is one reason why it has been the focus of attention
in the “theory of X” tradition. Matters are less straightforward, however,
when one considers both the full-blown capacities we have for folk psy-
chological explanation and some of our more advanced deployments of
folk psychology.
Consider full-blown folk psychology, which augments bare-bones folk
psychology with a heterogeneous bunch of further states and capacities
that the folk readily attribute to one another. Apart from the various emo-
tions, moods, and sensations already mentioned, full-blown folk psychol-
ogy includes character and temperament states (sturdy, reliable, happy-
go-lucky) and global cognitive assessments (rational, intelligent, scatty).
It is much less plausible to think that the realization of the capacity to
ascribe full-blown folk psychology is purely internal than to think so in
the case of bare-bones psychology. That is because these states have a
felt component, whether it be experiential or bodily (or both), and it is
dif¬cult to see how one could accurately and reliably ascribe such states
to others without knowing what they were like in one™s own case. Such
knowledge itself is procedural and has a bodily realization in that it in-
volves not simply having one™s brain in some internal state but, at least,
having one™s brain and body in a certain state.
The most obvious ploys for proponents of the theory view of folk psy-
chology would be to argue that (i) full-blown folk psychology can be re-
duced to bare-bones psychology, or (ii) however important experiential
and bodily aspects are to the acquisition of folk psychology, they do not
form part of its realization, which is purely internal. Both of these options
Thinking Through and Beyond the Body
208

seem unpromising, however, in light of the argument of Chapters 5
and 6: (i) would seem to involve either the sort of hierarchical decompo-

<<

. 33
( 62 .)



>>